How to Measure Anything

Doug Hubbard Expert The Advisory Council www.TACadvisory.com

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

How to Measure Anything

I conducted 55 major risk/return analysis projects so far that included a variety of impossible measurements I found such a high need for measuring difficult things that I decided I had to write a book The book will be released in July 2007 with the publisher John Wiley & Sons This is a sneak preview of many of the methods in the book

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

A Few Examples
Org. EPA Problem
Assess value of Safe Drinking Water Information System SDWIS Forecast fuel use for the battle field

Findings
Identified and measured previously unnoticed risks & benefits Found factors that better correlated to actual fuel use

Benefit
$15 million improved NPV from reprioritizing functions

US Marine Corp

$50 million savings per year by reducing unnecessary battlefield inventory $2 million improved net benefit reprioritizing rollout

American Express

Analyze ROI of a dedicated test network for new applications

Determined key metric to prioritize replacements was downtime reduction

Source: Hubbard Decision Research

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

A Few More Examples

Risk of IT The risk of obsolescence The value of a human life The value of saving an endangered species The value of public health The value of IQ points lost by children exposed to methyl-mercury The value of better security The future demand for space tourism The value of better information The value of information availability Productivity and performance

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

Three Measurement Muses

Eratosthenes measured the Earths circumference to within 1% accuracy Enrico Fermi the physicist who used Fermi Questions to break down any uncertain quantity (and was the first to estimate the yield of the first atom bomb)

Emily Rosa the 11 yr old who was published in JAMA (youngest author ever) for her experiment that debunked therapeutic touch
Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

Three Illusions of Intangibles

(The .com approach)
The perceived impossibility of measurement is an illusion caused by not understanding: the Concept of measurement the Object of measurement the Methods of measurement See my Everything Is Measurable article in CIO magazine
Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

Before We Measure We Ask:

Why do you want to know? How much do you know now? What is the value to additional information?

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

Why Do You Want To Know?

The EPA needed to compute the ROI of the Safe Drinking Water Information System (SDWIS) Why?: To prioritize three specific upgrades We built a spreadsheet model that connected the expected effects of the system to relevant impacts

Source: Hubbard Decision Research Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

Uncertainty, Risk & Measurement

Measuring Uncertainty, Risk and the Value of Information are closely related concepts, important measurements themselves, and precursors to most other measurements The Measurement Theory definition of measurement: A measurement is an observation that results in information (reduction of uncertainty) about a quantity. We model uncertainty statistically
ROI = 0% Risk of Negative ROI -25% 0% 25% 50% 75% Return on Investment (ROI) 100% 125%

Expected ROI

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

How Much Do You Know Now?

Most people can be taught how to subjectively assess their current state of uncertainty A small amount of training (three hours) can significantly improve calibration of estimates Comparisons of actual measures to original calibrated estimates show calibration works

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

10

Risk/ROI w/ Monte Carlo

Inputs
Administrative Cost Reduction
5% 10% 15%

% Improvement in Customer Retention

10% 20% 30%

Total Project Cost

$2 million $4 million $6 million

ROI
-50% 0% 50% 100%
11

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

What Is The Value of Measurement?

As a rule of thumb, the value of information is simply the cost of being wrong times the chance of being wrong The value of information on a range often just comes down to where the threshold is within the range.
B
Threshold: Below this point is losing money Relative Threshold (RT)=B/A

Worst Bound of the 90% CI; this is the undesirable end of the range

Best Bound of the 90% CI; this is the desirable end of the range

Source: How to Measure Anything: Finding the Value of Intangibles in Business

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

12

The IT(?) Measurement Inversion

Receives Most Attention

Costs

Least Relevant to Approval Decisions

Economic Relevance

Initial Development Costs Ongoing Maintenance/Training Costs

Receives Least Attention

See my article The IT Measurement Inversion in CIO magazine

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

Typical Attention

Benefits
A specific benefit (productivity, sales, etc.) Utilization (when usage starts and how quickly usage grows)

Chance of Cancellation
Most Relevant to Approval Decisions

13

Next Step: Observations

Now that we know what to measure and what its worth to measure, we can think of observations that would reduce uncertainty The value of the information limits what methods we should use, but we have a variety of methods available Take the Nike Method: Just Do It dont let imagined difficulties get in the way of starting observations
Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

14

The Power of Sampling Methods

Several clever sampling methods exist that can measure more with less data than you might think Examples: estimating the population of fish in the ocean, estimating the number of tanks created by the Germans in WW II, extremely small samples, etc.

Source: How to Measure Anything: Finding the Value of Intangibles in Business

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

15

The Math-less Statistics Table

Measurement is based on observation and most observations are just samples Reducing your uncertainty with random samples is not made intuitive in most statistics texts This table makes computing a 90% confidence interval easy
Source: How to Measure Anything: Finding the Value of Intangibles in Business

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

16

Measuring to the Threshold

Measurements have value usually because there is some point where the quantity makes a difference Its often much harder to ask How much is X than Is X enough
Source: How to Measure Anything: Finding the Value of Intangibles in Business

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

17

The Simplest Method

Bayesian methods in statistics use new information to update prior knowledge Bayesian methods can be even more elaborate that other statistical methods BUT It turns out that calibrated people are already mostly subjectively Bayesian

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

18

Improving on Human Judgment

The Lens Model is one method used to improve on expert intuition The chart shows the reduction in error from this method on intuitive estimates In every case, this method equaled or bettered the judgment of experts
Business failures using financial ratios Psychology course grades IQ scores using Rorschach tests Student ratings of teaching effectiveness Mental illness using personality tests Changes in stock prices Graduate students grades Life-insurance salesrep performance Cancer patient life-expectancy 0% 5% 10% 15% 20% 25% 30% 35% 40%

Reduction in Errors

Source: How to Measure Anything: Finding the Value of Intangibles in Business

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

19

Prediction Markets
Simulated trading markets are a proven method of generating probabilities for uncertain events Research shows that it works even without purely monetary reward systems

Source: Servan-Schreiber, et. al. Electronic Markets, v 14-3, September 2004

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

20

Applied Information Economics

The basic AIE model was used in several of these examples where intangibles dominated the decision Even though it has been used on problems as different as IT portfolio prioritization Forecasting fuel for the battlefield the same basic process applies
Start: Define Decision(s) Build a decision model (Ongoing evaluations decisions and course corrections)

Select/Calibrate estimators

Add/update values in model

Compute Information Values

Collect/Analyze New Observations (PseudoBayesian or additional expert polling)

Do any Variables have a significant EVPI? Yes

No

Make Decision

Select/Design Measurement Instrument

Decompose the Variable (if possible)

Source: How to Measure Anything: Finding the Value of Intangibles in Business

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

21

Final Tips
Learn how to think about uncertainty, risk and information value in a quantitative way Assume its been measured before You have more data than you think and you need less data than you think Methods that reduce your uncertainty are more economical than many managers assume Dont let exception anxiety cause you to avoid any observations at all Just do it
Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

22

More Information:
Several of my articles in CIO magazine, InformationWeek and other periodicals discuss measurement - especially the risk and value of IT. All of them are linked on www.hubbardresearch.com. The book will have its own web site at www.howtomeasureanything.com. The site will be online before the book is released in July. It will include detailed examples and spreadsheets mentioned in the book as well as discussion groups for readers. Contribute to The Measurement Challenge by sending us your examples of difficult measurements.

Copyright 2007 Hubbard Decision Research and The Advisory Council, Inc.

23